Use of blue dextran as a probe for the Nicotinamide Adenine

Jul 26, 1976 - and nicotinamide adenine dinucleotide(NAD) are reactants common ... phosphoryl donor for phosphate transfer reactions catalyzed by kina...
0 downloads 0 Views 1MB Size
Download PDF
St el 1wagen

92

Accounts of Chemical Research

Use of Blue Dextran as a Probe for the Nicotinamide Adenine Dinucleotide Domain in Proteins Earle Stellwagen Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242 Received July 26, 1976

Nucleotides such as adenosine 5’-triphosphate (ATP) and nicotinamide adenine dinucleotide (NAD) are reactants common to several hundred different biological reactions, each catalyzed by a different enzyme. The mononucleotide ATP commonly functions as the phosphoryl donor for phosphate transfer reactions catalyzed by kinases specific for a wide variety of phosphate acceptors, while the dinucleotide NAD functions as a reversible hydrogen acceptor in a variety of biological oxidative reactions catalyzed by dehydrogenases.

OH OH

a d e n o s i n e t r i p h o s p h a t e ( ATP)

OH OH

n i c o t i n a m i d e a d e n i n e d i n u c l e o t i d e ( NADt)

The recent outpouring of enzyme structures described to high resolution by x-ray crystallographers has led to the recognition that enzymes which catalyze reactions involving ATP or NAD frequently possess a common architectural feature called the NAD binding d0main.l This domain, illustrated schematically in Figure 1, involves a continuous 150-residue segment of a polypeptide chain arranged into a @-sheetcore composed of four to six parallel @ strands connected by helical intrastrand loops located above and below the p sheet. The amino acid residues surrounding the C terminus of the @ sheet form a high affinity binding site for a mono- or dinucleotide and dictate its selectivity. Crystallographic measurements have demonstrated that the NAD domain in certain dehydrogenases,lkinases,lt2 a phosphorylase,3and a mutase4 forms the binding sites for NAD, ATP, adenine ribose 5’-phosphate (AMP), and a sugar phosphate, respectively. Blue dextran is a high molecular weight (about 2 X lo6) water-soluble conjugate of dextran and a blue dye Earle Stellwagen obtained his Ph.D. degree in 1963 from the University of Cacfomia at Berkeley under the drection of Howard K. Schchman. After spending nearly 2 years as a National Institutes of Health (NIH) postdoctoral fellow with Hans Tuppy at the University of Vienna, he joined the University of Iowa where he is now Rofessor of Biochemisby. His research interest is in structure-function relationships of proteins.

commonly used to measure the void volume of exclusion chromatographic columns. Many investigators have inadvertently observed that blue dextran associates rather tightly with certain proteins when it is added to protein solutions prior to exclusion chromatography in solvents of low ionic strength. Among the earliest reported blue dextran-protein complexes are those involving pyruvate kinase5 and phosphofructokinase.6 Such complexation has subsequently been exploited to selectively enrich7 or deplete’ heterogeneous protein solutions of a desired component using either blue dextran. m ~ b i l e ~ ”or - ~imm~bilized~~-O ~’ Recent chromatographic7’and spectrophotometricg measurements indicate that the blue dye in blue dextran specifically complexes with protein ligand sites constructed by the NAD binding domain. This Account will review the evidence for the specificity of complexation of the blue dye of blue dextran with proteins and illustrate the broad utilization of this specificity for structural predictions and protein manipulations. Dye Specificity The blue dye in blue dextran is reactive blue 2, Color Index No. 61-211, commercially known as Cibacron Blue F3GA. Comparison of a space-filling model of reactive blue 2 with that of NAD oriented’ as it binds to the NAD binding domain of a dehydrogenase reveals distinct structural similarities. As shown in Figure 2, (1)M. G. Rossmann, A. Liljas, C. I. Branden, and E. J. Banaszak, Enzymes, 3rd Ed., 11, Chapter 2 (1975). ( 2 ) (a) T. N. Bryant, H. C. Watson, and P. L. Wendell, Nature (London), 247,14 (1974); (b) C. C. F. Blake and P. R. Evans, J. Mol. Biol., 84,585 (1974) ~~. . .,.

(3) R. J. Fletterick, J. Sygusch, H. Semple, and N. B. Madsen, J . Biol. Chem., 251, 6142 (1976). (4) J. W. Camubell, H. C. Watson, and G. I. Hodgson, Nature (London), 250, 301 (1974). (5) R. Haeckel, B. Hess, W. Lauterborn, and K. H. Wuster, HoppeSeyler’s Z. Physiol. Chem., 349, 669 (1968). (6) G. Kopperschlager, G. Freyer, W. Diezel, and E. Hofmann, FEBS Lett., 1, 137 (1968). (7) (a) A. C. W. Swart and H. C. Hemker, Biochim. Biophys. Acta, 222, 692 (1970); (b) G. E. J. Stahl, J. F. Koster, H. Kamp, L. Van Milligen-Boersma, and C. Veeger, ibid., 227, 86 (1971); ( c ) K. G. Blume, R. W. Hoffbauer, D. Busch, H. Arnold, and G. W. Lohr, ibid.,227,364 (1971); (d) H. White and W. P. Jencks, J . €302. Chem., 251,1708 (1976); ( e ) G. Komerschlazer. W. Diezel. R. Frever. S. Liebe, and E. Hoffmann. Eur. J. btoehem.,-22; 40 (1971); (f) R: L. Easterday and I. M. Easterday, “Immobilized Biochemicals and Affinity Chromatography”,R. B. Dunlap, Ed., Plenum, New York, N.Y., 1974, p 123; (g) P. Roschlau and B. Hess, Hoppe-Seyler’s 2. Physiol. Chem., 353,441 (1972); (h) L. D. Ryan and C. S. Vestling, Arch. Biochern. Biophys., 160,279 (1974); (i) S. T. Thompson, K. H. Cam, and E. Stellwagen,Proc. Natl. Acad. Sci. U.S.A., 72,669 (1975); (i) L. P. Solomonson, Plant Physiol., 56, 853 (1975); (k) J. J. Witt and R. Roskoski, Jr., Biochemistry, 14, 4503 (1975); (1) R. Roskoski, Jr., C. T. Lim, and L. M. Roskoski, ibid., 14,5105 (1975); (m) R. Kobayashi and V. S. Fang, Biochem. Biophys. Res. Cornrnun., 69,1080 (1976); (n) L. Vician and G. H. Tishkoff, Biochim. Biophys. Acta, 434, 199 (1976); (0)W. J. Jankowski, W. von Muenchhausen, E. Sulkowski, and W. A. Carter, Biochemistry, 15, 5182 (1976); (p) E. Stellwagen and B. Baker, Nature (London),261,719 (1976); (9) J. G. Moe and D. Piszkiewicz, FEBS Lett., in press. (8) J. Travis and R. Pannell, Clin. Chirn. Acta, 49, 49 (1973). (9) S. T. Thompson and E. Stellwagen, Proc. Natl. Acad. Sci. U.S.A., 73, 361 (1976).

NAD Domain in Proteins

Vol. 10, 1977

Figure 1. A diagrammatic representation of the NAD binding domain of lactate dehydroganse.'O The parallel p strands are labeled PA. . .& while the a helices are labeled ag. . .a3G. The position of the' structural elements of NAD, adenine, ribose, phosphate, and nicotinamide, denoted as A, R, P, and N, respectively, are indicated as they bind to the C terminus of the 0 sheet of the NAD binding domain.

reactive blue 2

\

S020Na

'cI

the anthraquinone and terminal phenylsulfonate rings of the dye can be oriented as the adenine and nicotinamide rings of NAD, respectively. The two internal rings of the dye mimic the two ribosofuranose rings of NAD, and two of the dye anionic sulfonate groups are positioned similarly to the two anionic phosphates of NAD. The structure of reactive blue 2 is also similar to the structures of ATP and coenzyme A as shown in Figure 2. The interaction of reactive blue 2 with the M4 isozyme of lactate dehydrogenase, whose four NAD binding sites are constructed by an NAD binding domain,1° has been studiedgin detail. Equilibrium dialysis measurements demonstrate that four dye molecules bind independently to tetrameric lactate dehydrogenase, each with an association constant of 1.4 X lo6M. Double-reciprocal plots of the dependence of the rate of catalysis on the concentration of NADH in the presence and absence of dye indicate that the dye is a competitive inhibitor with NADH. The reciprocal of the competitive inhibition constant, 7.7 f 0.6 X lo6 M, compares favorably with the association constant measured by equilibrium dialysis. Thus the single dye-binding site on each subunit must also be the NAD site constructed by the NAD binding domain. This conclusion is supported by difference spectral measurements of the dye-enzyme complex in the presence of ligands specific for the NAD binding domain. If the anthraquinone and terminal phenyl(10)J. J. Holbrook, A. Liljas, s. J. Steindel, and M. G. Rossmann, Enzymes, 3rd Ed., 11, Chapter 4 (1975).

93

sulfonate rings of the dye are inserted into the apolar pockets in the NAD binding domain designed to bind the adenine and nicotinamide rings of NAD, then complexation of the dye with the enzyme should cause a red shift in the absorption spectrum of the dye. This red shift does occur, producing a difference spectrum having a positive maximum at 660 nm, as shown in the difference spectra illustrated in Figure 3A. Both the occurrence of the isosbestic point at 584 nm (Figure 3A) and the hyperbolic dependence of the difference absorbance at 660 nm on the total dye concentration (Figure 3B) indicate formation of a discrete dye-enzyme complex. An association constant of 2.3 f 0.7 X lo6 M calculated from this hyperbolic dependence is in good agreement with the association constant for the dyeenzyme complex calculated from equilibrium dialysis and kinetic measurements. Addition of increasing concentrations of the ligands NAD or NADH to the dye-enzyme complex causes a progressive decrease in the difference absorbance maximum, as shown in Figure 3C. Most significantly, the ligand which binds most tightly to the NAD binding domain, the reduced form of NAD (NADH), is most effective in reducing the difference absorbance. Association constants for the NADH-enzyme and NAD-enzyme complexes calculated from the values shown in Figure 3C, 1.7 f 0.7 X lo6 M and 2.3 i 1.4 X lo3 M, respectively, are in good agreement with association constants previous€y measuredlo for these complexes. The ability of relatively high concentrations (>0.2 M) of the nonspecific ligand NaCl to dissociate the dye-enzyme complex indicates that at least a portion of the interaction between the dye and the NAD binding domain is electrostatic as suggested above. Similar resultsg were obtained in studying the interaction between ATP and the dye with the NAD binding domain of phosphoglycerate kinase which forms the ATP binding site in this enzyme.2 Catalytic measurements indicate that the dye is a competitive inhibitor for ATP. Addition of the dye to the enzyme produces difference spectra characteristic for insertion of the dye into the adenine pocket to form a discrete dye-enzyme complex, while addition of ATP to the complex indicates that the natural ligand displaces the dye from the NAD binding domain. The association constant for the ATP-enzyme complex falls within the range of values measured using other procedures.ll The specific complexation of the dye with the NAD binding domain of a variety of proteins was also investigated using immobilized blue dextran covalently linked through at least one dye moiety of each dextran to Sepharose. Thus the remaining dye molecules attached to each dextran are in principle available for interaction with proteins. Such interaction with the NAD binding domain is feasible since the covalent linkage of dextran to the triazine ring of the dye occurs on the side of the dye opposite to that considered (Figure 2) to contact the domain. As shown in Table I, each of the eight proteins known from crystallographic meas~rementsl-~ to contain an NAD binding domain specifically binds to blue dextran-Sepharose chromatographic columns. By contrast, proteins known not to contain an NAD binding domain12or to contain (11)(a) W.K. G. Krietsch and T. Bucher, Eur. J.Biochem., 17,568 (1970);(b) M.Larsson-Raznikiewicz,Arch. Biochem. Biophys., 158,754 (1973).

Stellwagen

94

Accounts of Chemical Research

Figure 2. Comparison of a Corey-Pauling-Koltun (CPK) model of reactive blue 2 and various nucleoside phosphates. The models represent, from left to right, the NAD-pyruvate adduct, NAD, reactive blue 2, coenzyme A, and ATP. X . nrr

Sepharose I

"

0

Aso20Nn

1

-

/Ih

I E

0

bW

a

12 % domains either without an adenine poc%et13or having their secondary structural elements connected in a different order14 do not bind to these columns. Each of the eight proteins containing an NAD binding domain is most readily eluted from blue (12) (a) M. F. Perutz, H. Muirhead, J. M. Cox and L. C. G. Goaman, Nature (London),219, 131 (1968); (b) P. M. Colman, I. N. Jansonius, and B. W. Matthews, J. Mol.Biol., 70,701 (1972); (c) J. S. Richardson, K. A. Thomas. B. H. Rubin. and D. C. Richardson. Proc. Natl. Acad. Sei. U.S.A., 72, 1349 (1975). (13) (a) A. Arnone, C. J. Bier, F. A. Cotton, E. E. Hazen, Jr., D. C. Richardson, and J. S. Richardson, Proc. Natl. Acad. Sci. U.S.A., 64,420 (1969); (b) C. S. Wright, R. A. Alden, and J. Kraut, Nature (London), 221,235 (1969); (c) J. Drenth, W. G. J. Hol, J. N. Jansonius, and R. Koekoek, Cold Spring Harbor Symp. Quant. Biol., 35, 107 (1971); (d) K. D. Watenpaugh, L. C. Sicker, L. H. Jensen, T. LeGall, and M. Dubourdieu, Proc. Nutl. Acad. Sci. U.S.A., 69,3185 (1972); (e) R. D. Andersen, P. A. Apgar, R. M., Burnett, G. D. Darling, M. E. LeQuesne, S. G. Mayhew, and M. L. Ludwlg, ibzd., 69, 3189 (1972). (14) (a) R. J. Fletterick, D. J. Bates, and T. A. Steitz, Proc. Natl. Acad. Sci. U.S.A., 72,38 (1975); (b) D. W. Banner, A. C. Bloomer, G. A. Petsko, D. C. Phillips, C. I. Pogson, and T. A. Wilson, Nature (London),255,609 v

(1975).

Figure 3. Spectral measurements of reactive blue 2/M4-lactate dehydrogenase mixtures. (A) Difference spectra of the dye in the presence of increasing concentrations of the enzyme. The sample cuvette initially contained 12 pM enzyme subunits while both the sample and reference cuvettes contained (a) 2.15 pM, (b) 4.28 pM, and (c) 7.47 pM dye. (B) Difference extinction at 660 nm of the dye, expressed in enzyme subunit concentration units, as a function of the total dye concentration present. The sample cuvette initially contained 12 pM enzyme subunits while the indicated total concentrations of dye were added to both the sample and reference cuvettes. (C) The effect of increasing concentrations of the ligands NAD (A) or NADH ( 0 ) on the difference extinction of a dye:enzyme mixture measured at 660 nm. The sample cuvette initially contained 11.6 p M enzyme subunits and 26.1 pM dye only, The indicated concentrations of NAD or NADH were added to both cuvettes. The solvent for all these spectral measurements was 10 mM Tris-€IC1buffer, pH 7.5, containing 0.5 mM 2-mercaptoethanol.

dextran-Sepharose columns by solvents containing the nucleoside or sugar phosphate ligand which binds most tightly to the NAD binding domain of the protein being

NAD Domain in Proteins

Vol. 10, 1977

95

Table I Interaction of Proteins with Blue Dextran-Sepharose Specific interaction Proteins of known tertiary structure Lactate dehydrogenase: Malate dehydr~genase~l Alcohol d e h v d r o ~ e n a s ef *~i Glyceraldehyde-ihosphated e h y d r ~ g e n a s e ~ ~ Adenylate kinase7l Phosphoglycerate kinase' Phosphoglucomutase7 Phosphorylase a' Proteins of unknown tertiary structure Pyruvate kinase7f ~ l Phosphofructokinase7e-i Protein kinase?k , m Adenylate cyclase7p DNA polymerase 171 Phosphogluconate. dehydrogenase?f * i Nitrate reductase7] Aldolase' Fructose bi~phosphatase~' Choline acetyltransferase'' Isoleucyl-tRNA synthetase7 CAMP phosphodiesterase'8a Methylmalonyl-CoA isomerase' ' Citrate synthetaselsd

Nonspecific interaction Ribonuclease?f*i Chymotrypsin0 e n 7 f Cvtochrome c 7fi

Hemoglobin7 Thermolysin SuDeroxide dismutaselsb Micrococcal nuclease7 Subtilisin7l Apoflav~doxin~~ Hexokinase' Triosephosphate isomerase'

Serum albumin? ** Enolase' Interferon7O Clotting factor X7"

Ovalbumin7f Thyroglobulin7 Aspartate transcarbamyltye Dihydrofolate reductase'l Phosphoglucomutase' Phosphoglucoisomerase '

examined. In all cases, the ionic strength of ligand solutions required to rapidly elute each protein is at least an order of magnitude less than that of NaCl solutions which effect a similar rapid elution of protein. These results suggest that complexation of the dye with the NAD binding domain involves apolar, electrostatic, and perhaps hydrogen-bond interactions similar to those involving complexation of the specific natural ligand. As shown in Table I, three basic proteins not to contain an NAD binding domain, ribonuclease, chymotrypsinogen,and cytochrome c, are also retained by the anionic blue dextran-Sepharose columns. However, each of these proteins is eluted by a concentration of NaCl equivalent in ionic strength to the concentration of specific ligand. Thus, specific complexation of the immobilized dye with proteins containing the NAD binding domain can be distinguished from ion-exchange chromatography of the anionic dye with cationic proteins by comparing the effectiveness of solutions of specific and nonspecific ligands of comparable ionic strength in eluting bound proteins. Immobilized dye-Sepharose columns have also been prepared using reactive blue 2 and its isomer, reactive blue 5 (1). Phosphofructokinase has been reported to Q

No interaction

"2

N-3 CI 1

columns but not to reactive blue 5-dextran-Sepharose columns.l8 Since ATP bothI6 elutes the enzyme from immobilized reactive blue 2 columns and competes with mobile reactive blue 2 for the allosteric binding site, it is very likely that the allosteric site on phosphofructokinase is constructed by an NAD binding domain. The failure of reactive blue 5 to bind to this site suggests that the interaction between reactive blue 2 and an NAD binding domain requires a precise orientation of apolar and ionic groups and is not the result of a rather nonspecific apolar and/or ionic interaction of a portion of the dye with enzyme surfaces.

Structural Predictions Since blue dextran-Sepharose functions as an affinity column for ligand sites constructed by the NAD binding domain, it should be possible to predict the occurrence of this domain in the structures of proteins which have not yet been subjected to crystallographic analyses. A protein containing an NAD binding domain would be expected to bind to the column and be eluted by a solvent containing the ligand specific for the protein having an ionic strength at least an order of magnitude less than the NaCl solvents which produce a comparable elution of the protein. Using these criteria, rabbit muscle phosphorylase a was predicted to contain the NAD binding domain while yeast hexokinase was predicted not to contain the NAD binding domain.17 Subsequent crystallographic measurements3J4 have substantiated both of these predictions. It should be relatively simple to predict the presence of the NAD binding domain in a wide range of proteins. All that is required in principle is a small blue dextran-Sepharose chromatographic column, a crude homogenate containing the protein of interest and a specific assay procedure for the biological function of the protein. As shown in Table I, a variety of enzymes

bind specifically to reactive blue 2-dextran-Sepharose (15)(a) F. M. Richards and H. W. Wyckoff, Enzymes, 3rd Ed., 4, Chapter 24 (1971); (b) J. Kraut, ibid., 3, Chapter 5 (1971); (c) R. E. Dickerson and R. Timkovich, ibid., 11, Chapter 7 (1975).

(16)H.J. Bohme, G. Kopperschlker, J. Schultz, and E. Hofmann, J. Chromatogr., 69,209 (1972). (17)E. Stellwagen, R.H. C m , S. T. Thompson, and M. Woody, Nature (London),257, 716 (1975).

46

Stellwagen

which Lind A'TP, siach as pyruvate kinase, protein kinase, adenylate cyclase, and DNA polymerase I, appear to contain the NAD binding domain, while aspartate transcarbsmylase does not. Similarly, some ernzyrtlrs which bind NAD and/or NADP, such as nitrate redut:tase and phosphogluconate dehydrogenase, appear to contain the NAD binding domain, while dlhydrofolale reductase does not. Several enzymes spwific for coenzyme A or acyl coenzyme A, including citrar e synthetase, choline acetyltransferase, and methylrilalonyl-CoA isomerase, bind specifically to blue dextran Sepharose columns, suggesting that the coenzyme A sites can also be constructed by the NAD birding domain. The binding of fructose bisphosplratase to blue dextran-Sepharose columns and its specific elution with the effc~torV-AMP suggest that allosteric as well as substrate sites may also be construcied by the NAD binding domain. Finally, the specific interaction of aldolase with blue dextranSqihai o w sriggests that enzymes other than phos~~~~~~~~~~~~~~~~~e may have sugar phosphate binding sites cotiatructed by an NAD binding domain. 'I'hest: y.csdkc,suggest that a variety of proteins having a wide r m g e of biological functions appear to contain the NAr) binding domain. This chromatographic procedure can he used to systematically search for proteins c oniaining t h e NAD binding domain in order to ctjnfitr\lct evol i~tionaryrelationships, It would be of ~d~tirsulcar irlterest. t o survey the enzymes involved in nucleic acid, plulein, a d fatty acid synthesis for the prct;ence of the N A binding domain. Such olutiqmmry relationships can be extended if the aknirao acid sequences of proteins which specificdy bind to bluc dextran-Sephnrose are known. The sequences of sua'h proteins call be ana1yzedlg for contiritmus regions likely to form the alternating pattern t d 0 strarrds a i d 01 helices characteristic of an NAD ioiidirrg doniairu The sequence of such regions can then Ire gt:~ieiidlycompared with the sequence forming the NAD binding ctornain in a model protein such as lactate dehydrogenase using niinirnum base change per amino aaaid codon ci iQcria.l Such analysis of the sequence of aldolase predicts2' that residues 146- 300 can form secondsy structural elements characteristic of the N bindhrg domain and are genetically related to the sequence of laclate dehydrogenase which forms its NAD binding aomaira. of Bites and A nc~vcluse of blue dextran-Sepharose chromatography has been employed in the resolution of CAMPdependent protein kinase.7k 'B'bis protein consists of a catalytic submiit specific for the phosphate donor ATP and n regiwla.ta~rysubunit which masks the ATP site on the catalytic subunit. Addition of 3',5'-cyclic AMP (CAMP)to lhe onzyzme remlts in the dissociation of the regulatory and cx~talyticsubunits, thereby exposing the ATP binding site on the catalytic subunit. The holoe,.ezyare consisting of both kinds of subunits does plot bine! to blue dextranSepharose columns. In the presence of CAMP,the dissociated catdybic subunit can thwi he eluted from the column using solvents con~~~~~~~~

(18)(0) S T 'I'hompson and E Steilwagen, in preparation; (b) R. Ledyzi , (CJ 91. Tkachuck; ;uid (a)P Srere, all personal communications. (19) P. Y Cliuu and (d. D Fasman, Biochemistry, 11, 222 (1974). (20) F' Stellwagen, J Mol Biol., 1108, 903 (1976)

Accounts of Chemical Research

taining either 1 mM ATP or 0.2 M KCI. These results indicate that blue dextran-Sepharose chromatography can be used e0 determine the exposure of ligand sites constructed by an NAD binding domain provided the affinity of the ligand site for a potential masking agent is greater than the affinity of the site for the dye. Preliminary measurements with ribosomes and piatelets suggest that these assemblies contain exposed sites constructed by NAD binding domains. The results obtained with protein kinase also demonstrate that blue dextran-Sepharose chromatography can be used to fractionate protein complexes and subcellular assemblies provided one or more of the protein substituents contains an exposed ligand site constructed by an NAD binding domain and that conditions can be found for dissociation of the substituents while retaining their functional integrity. Ligand Exchange or Removal A variety of proteins such as glyceraldehydephosphate dehydrogenase21and phosphofructokinasez2 have such a strong affinity for their ligands that they noranally occur as protein-ligand complexes. In order to examine aspects of the structure and function of such proteins, it is often desirable either to prepare the apoenzyme and/or to replace the ligand with an analogue. Tightly bound nucleoside phosphate ligands are usually removed by repeated addition of activated charcoal to protein solutions, a procedure noted for low recovery and possible damage to the a p ~ p r o t e i n ,By ~~ contrast, recent measurement^^^ indicate that ligands tightly bound to the NAD binding domain of proteins can be easily removed using blue dextran-Sepharose chromatography resulting in the quantitative recovery of both the bound ligand and the apoprotein. Application of a solution of glyceraldehyde-phosphate dehydrogenase to a small blue dextran-Sepharose column equilibrated with 10 mM Tris-HC1 buffer, pH 7.5, results in the retention of the protein and elution of all the formerly protein-bound NAD. The apoenzyme is then eluted by the Tris buffer containing 1 M NaC1, resulting in recovery of at least 95% of the initial enzymic activity. The success of this procedure depends on the relative affinity of the NAD binding domain of the apoprotein for the immobilized dye and for the bound ligand and on the much greater concentration of immobilized dye relative to bound ligand. Since sites specific for a variety of nucleotide ligands appear to be constructed by NAD binding domains, this procedure should be useful both for preparation of apoproteins from holoproteins and for measuring the molar ratio of ligand and apoprotein in the holoprotein. Alternatively, the natural ligand can be directly replaced with an analogue by elution of the column-bound apoenzyme with so!utions containing the desired analogue. Complexation of apoproteins with spin-labeled, 13C-enriched or fluorescent ligand analogues should be achieved in (21) S. F. Velick, J. P. Baggott, and J. M. Sturtevant, Biochemistry, 10, 779 (1971). (22) K. H. Cass and E. Stellwagen, Arch. Biochem. Biophys., 171,682 (1975). (23) (a) J. F. Taylor, S. F. Velick, G. T. Cori, C. F. Cori, and M. W. Stein, J. Biol. Chem., 173, 619 (1948); (b) W. Ferdinand, Biochem. J., 92, 578 (1964). (24) S. T. Thompaon, R. H. Cam, and E. Stellwagen, Anal. Biochem., 72, 293 (1976).

Vol. 10, 1977

NAD Domain in Proteins

excellent yields by these procedures. Protein Purification Mobile blue dextran was first used in protein purification as a final step in the purification of pyruvate k i n a ~ e . Blue ~ ~ ' ~dextran was added to a heterogeneous pyruvate kinase solution and the mixture applied to a Sephadex G-200 exclusion column equilibrated and eluted with a solvent of low ionic strength. The large proteins and the pyruvate kinase-blue dextran complex were eluted in the void volume in advance of the smaller proteins. Reapplication of the void volume eluate to a second Sephadex G-200 exclusion column equilibrated and eluted with a solvent of high ionic strength resulted in the dissociation of the blue dextran-pyruvate kinase complex and the separation of the pyruvate kinase from the large proteins which continued to be totally excluded. Immobilization of the blue dextran by covalent attachment to an inert support together with recognition of the specificity of the dye for the NAD binding domain has greatly expanded the utility of blue dextran-protein complexation for protein purification. Both the blue dextran and the dye itself have been covalently attached to Sephadex, Sepharose, and polyacrylamide gels and to c e l l ~ l o s e . ~Comparison "~~~~~~~ of the binding and elution properties of pyruvate kinase, creatine kinase, and phosphofructokinase with such columns demonstrated that covalent attachment of blue dextran to Sepharose 4B or 6B is most versatile for chromatographic fractionation of p r ~ t e i n s . ~Appli~J~ cation of the supernatant from a homogenate of Escherichia coli to blue dextran-Sepharose and elution with an ATP gradient result in about a 1500-fold purification of phosphofructokinase in good yield, giving an enzyme solution having the specific activity and gel electrophoretic pattern characteristic of the purified purification to homoe n ~ y m eA . ~similar ~ ~ ~ one-step ~ geneity using blue dextran-Sepharose chromatography has been achieved for nitrate reductase from chlorella or spinach using an NADH gradient to selectively elute the enzyme.7i A variety of enzymes of markedly enhanced enrichment can be recovered from a single homogenate by elution with successive concentrations of gradients of different ligands.7f Preliminary measurements with purified lactate dehydrogenase indicate that the interaction of this enzyme with blue dextran-Sepharose chromatographic columns is not perturbed by the presence of nonionic detergents. This result suggested that blue dextranSepharose chromatography could be useful for the purification of membrane bound proteins containing the NAD binding domain which are solubilized by nonionic detergents. The chromatographic profiie for elution of Lubrol-solubilized adenylate cyclase from bovine brain is illustrated in Figure 4. This chromatographic procedure results in over a 100-fold enhancement of adenylate cyclase activity in good ~ i e l d . ~The P purification of NADH-cytochrome b5 reductase solubilized from rabbit liver microsomes with Triton X-100 can also be achieved using blue dextran-Sepharose chromatography.lab Blue dextran-Sepharose chromatography of proteins offers several advantages over affinity chromatography using immobilized ligand columns. Blue dextranSepharose is relatively inexpensive and easily made

97

Fraction Number

Figure 4. Blue dextran-Sepharose chromatography of bovine brain adenylate cyclase. Frozen bovine brain cortex was homogenized and the insoluble material extracted with 2% 1,ubrol PX. The Lubrol extract was applied to a 9-mL column of blue dextran-Sepharose equilibrated with 10 mM Tris-HC1buffer, pH 7.5, containing 5 mM NaF and 1 mM EDTA. The column was washed with the equilibration solvent and then eluted with a h e w ATP gradient contained in 100 mL of the equilibration solvent as shown. All of the adenylate cyclase activity bound to the column eluted with the central peak in the elution profile, giving a 150-fold enzyme purification to near homogeneity.

from materials commonly found in research iaboratories. The blue chromophore is not biodegsadabla md thus is stable to the presence of catabolic enzymes In crude cellular extracts such as phosphatases which metabolize nucleoside phosphate affinity columns. Most importantly, a single column can be utilized to selectively enhance the relative concentration of a wide range of proteins, as illustrated in Table I. This selectivity is achieved by the choice and concentration of the mobile ligand used to elute the column containing the bound protein(s) of interest and i s based on the relative affinity of proteins for the immobilized dye and for the mobile ligand and the relative concentrations of the immobilized dye and the mobile ligand.

Investigation of Catalytic Sites As noted above, reactive blue 2 appears to specififically bind to the NAD binding domain in each subunit sf lactate dehydrogenase. The association c o n s h t for the enzyme subunit-dye complexg is about an order of magnitude greater than the association constant for the subunit-NADH complex and three orders of magnitude greater than the value for the subunit--NADcomplex.1° The greater affinity of the dye is surprising since the dye contains an anionic sulfonate group on its terminal phenyl ring while nicotinamide contains no anionic group. The association constant of each enzyme subunit for the covalent NAD-pyruvate adduct, which has an anionic group on the nicotinamide moiety (Figure l), is also greaterz5 than that for NADH or NAD. Crystallographic measurements1° of the enzyme NAD-pyruvate complex indicate that the structure of this binary complex is isomorphous with the enzymeNAD-oxalate ternary complex and is distinct from the enzyme-NAD binary complex. These comparisons suggest that the dye may span the coenzyme and substrate binding sites and that the enzymedye binary complex may be considered as a ternary complex analogue. While lactate dehydrogenase cannot be eluted from blue dextran-Sepharose columns by solvents containing either 15 mM NAD or 10 rnM pyruvate, the enzyme can be eluted by the same solvents (25)

L.J. Arnold, Jr., and N. 0.Kaplan, J. Biol. Chern.,249,662 (1974).

St e 11wage n

98

0

5 IO Bed Volumes

Figure 5. Blue dextran-Sepharose chromatography of yeast phosphoglycerate kinase. Fifty micrograms of the purified enzyme was added to a series of columns each containing 1 mL of blue dextran-Sepharose equilibrated with 10 mM Tris-HC1buffer, 0.5 mM 2-mercaptoethanol, pH 7.5. Each column was washed with five bed volumes of the equilibration solvent, as shown, and then with a t least five bed volumes of the equilibration solvent containing the indicated ligand concentrations. (A) concentration of ATP present in the eluting solvents: (0) 0.2 mM, ( 0 )0.4 mM, (A) 0.7 mM, and (A) 1.0 mM. (B) Concentration of 3-phosphoglycerate (PGA) present in the elution solvents: (0) 0.1-1.0 mM and ( 0 )2 mM. (C) Concentration of both ATP and PGA present in the elution solvents: (0) 0.1 mM, (A)0.2 mM, and ( 0 )0.4 mM.

containing only 1 mM NAD and 1 mM pyruvate.26 Therefore the coenzyme and substrate together are much more effective in displacing the enzyme from the dye than either individually, as would be expected if the dye occupied the binding sites for both the coenzyme and the substrate. Similar observations pertain (26) B. Nadel-Girard and C . L. Markert, “Isozymes”, Vol. 2, C . L. Markert, Ed., Academic Press, New York, N.Y., 1975, p 45.

Accounts of Chemical Research

to kinase-dye binary complexes. The affinity of phosphoglycerate kinase for the dye is about two orders of magnitude greatergthan for either substrate ATP or 3-pho~phog1ycerate.l~ As shown in Figure 5, the enzyme is eluted from blue dextran-Sepharose columns much more easily using a combination of both substrates than either substrate individually. These resuh &o suggest that the enzyme-dye binary complex c m be considered as a ternary complex analogue of certain enzymes. The stability, specificity, and chromophoric properties of the dye should be particularly useful in probing the geometry of catalytic sites and the conformational changes resulting from the complexation of the dye at such sites.

Concluding Remarks The experimental results obtained using both mobile and immobilized reactive blue 2 me consistent with the view that the dye has a strong affmity for protein ligand binding sites constructed by an NAD binding domain. This affinity can be utilized to predict the presence of the domain in individual proteins contained in heterogeneous mixture, to determine the exposure of the domain in protein complexes and subcellular assemblies, to greatly simplify the purification of proteins containing the domain, to remove or exchange ligands bound to the domain, and to locate ligand binding sites adjacent to the domain. The growing awareness of the wide distribution of the NAD binding domain among proteins important to DNA and protein biosynthesis, energy production, and endocrine regulation affords great opportunity to utilize blue dextran for protein purification and structural analysis. This work was supported by US. Public Health Service Research Grants GM-13215 and GM-22109, both from the Institute of General Medical Sciences.